The laboratory of Michael Yarmolinsky has continued its study of the bacterial equivalent of mitosis. In eukarotes, errors in mitosis are associated with developmental abnormalities and cancer. In prokaryotes the process has remained ill understood despite the apparent simplicity of the known components. To study the process the Yarmolinsky laboratory is making use of the simplest genetic elements that are equitably distributed to daughter cells at cell division, namely plasmids of low copy number such as prophage P1. Plasmid-encoded participants in this partitioning process are few: two proteins, ParA and ParB, and a DNA site, parS, the plasmid centromere. Interactions of wild type and mutant components of the partition apparatus have been studied in vivo by assays of (1) plasmid retention under non-selective conditions, (2) the expression of genes at various distances from parS, including reporter genes fused to genes of the par operon itself (3) immunoprecipitated DNA reversibly cross-linked to protein, (4) plasmid topoisomer distributions as displayed in one- and two-dimensional gels and (5) fractionation of cell components. In the period since the last report, Yarmolinsky's group has concentrated on two questions: the biological significance of a remarkable spreading of ParB along DNA flanking the plasmid centromere (a phenomenon reported in 1999 by members of the laboratory) and the contribution of ParA to the partitioning process. Additionally, Yarmolinsky, in collaboration with European colleagues trained in his laboratory (and with colleagues at the University of Wisconsin), has devoted considerable effort to the preparation of an analysis of the P1 genome. Accomplishments include: (1). A demonstration that the normally extensive polymerization of ParB along DNA flanking parS contributes little, if at all, to partitioning, despite suggestive genetic evidence to the contrary. In the partitioning process, the potential to polymerize may be more significant than polymerization itself. (Rodionov and Yarmolinsky, MS in preparation). (2). The isolation of host mutants that affect the gene-silencing efficiency of ParB and that appear not to be in the structural genes of the only known host component in partitioning, the heterodimeric architectural protein, integration host factor. Mutations that affect gene silencing offer a novel approach to the identification of host factors affecting partitioning. (3). Evidence for the stable binding of ParA to the ParB-parS partition complex. The binding is such as to prevent free rotation of the plasmid DNA, is likely to be to a cellular structure, and is possibly made up of ParA itself.